An Efficient Approximate Algorithm for Winner Determination in Combinatorial Auctions
|
|
- Ophelia Cannon
- 6 years ago
- Views:
Transcription
1 An Efficient Approximate Algorithm for Winner Determination in Combinatorial Auctions Yuko Sakurai, Makoto Yokoo, and Koji Kamei NTT Communication Science Laboratories, 2-4 Hikaridai, Seika-cho, Soraku-gun, Kyoto, , Japan Introduction Auctions on the Internet have become especially popular in Electronic Commerce. Amongthe various studies on the Internet auctions, those on combinatorial auctions have lately attracted considerable attention. Combinatorial auctions simultaneously sell multiple items with interdependent values and allow the bidders to bid on any combination of items. Therefore, they tend to increase the buyers utilities and the seller s revenue. On the other hand, determiningwinners in combinatorial auctions intended to maximize the sum of the bids with disjoint sets of items is a complicated constraint optimization problem and shown to be NP-complete [4]. This problem has recently attracted the interest of the AI community as an application of search techniques [,5]. Although the previous methods have significantly improved the efficiency of the optimal winner determination algorithms, to solve large-scale problems, we eventually need to give up the idea of achieving the optimality of the obtained allocation and try to find a semi-optimal solution within a reasonable amount of time. In this paper, we introduce limited discrepancy search (LDS) techniques [2] to limit the search efforts to the part where good solutions are likely to exist. Experimental evaluations using various problem settings show that our algorithm finds the allocations that are very close to the optimal solutions (better than 95%) very quickly (in about % of the runningtime) and can be extended to large-scale problem instances. 2 Preliminaries 2. Winner Determination Problem A winner determination problem can be formulated as follows [4,5]. Let G = {, 2,,m} be the set of items to be auctioned and A be a set of bidders. Suppose a bidder i A can submit any bid b i (S) for any combination of items S G. Let b(s) be the the maximal bid for the bundle S, thatis, b(s) = R. Dechter (Ed.): CP 2000, LNCS 894, pp , c Springer-Verlag Berlin Heidelberg 2000
2 550Yuko Sakurai et al. Fig.. Search Tree Fig. 2. Discrepancies of Nodes max i A b i (S), and χ be the set of possible allocations, that is, χ = {S G S S = for every S, S χ}. Then, the goal is to find an allocation χ such that argmax χ b(s). () S χ 2.2 Problem Formalization as Search Problem We basically follow the problem formalization used in [5]. An example of a search tree is presented in Figure. Each node is associated with a bid, and a path from the root node to any other node represents one (partial) allocation, which consists of a sequence of disjoint bids. The child nodes of each node are limited to the nodes that include the item with the smallest index amongthe items that are not yet on the path but that do not include items that are already on the path. In [5], a search algorithm called IDA* [3] is used for searchingthis tree. We introduce a special data structure called bins [] to find quickly the children of each nodes. More specifically, we sort bids into bins, where a bin B j contains all bids where j is the smallest index amongthe items in a bid. The followingheuristic function h is used to estimate the possible maximal revenue for the items not yet allocated on the path in [5]. h = k {unallocated items} r(k), where r(k) = b(s) max {S;k S} S (2) If we re-calculate h in IDA* whenever a bid is appended to the path, h becomes more accurate, but this re-calculation requires significant overhead. We found that this re-calculation is necessary in IDA* to decrease the number of visited nodes. On the other hand, we can avoid this re-calculation since LDS is less sensitive to the accuracy of heuristic evaluations. We use an evaluation function f(n) = g(n) + h(n) to estimate the best (highest) price of the solutions obtained from node n. g(n) is calculated as the sum of the bids on the path appended to n. h(n) is a heuristic function defined as (2). We use f(n) used for orderingsiblingnodes and for pruning.
3 An Efficient Approximate Algorithm for Winner Determination 55. If stack is empty, then terminate the algorithm, otherwise, pop list from stack. 2. If list is empty, go to, otherwise, set n c to the first node in list, andremoven c. 3. If dis(n c) >D max, thengoto. 4. If f(n c) f max, thengoto. 5. If n c is a leaf node, then record the current path as a best solution, set f max to f(n c), set n c to the first node in list, removen c from list, and go to Expand n c. push list to stack, setn c to the best child, and set list to the rest of the children sorted by f, goto3. Fig. 3. Pseudo Code of LDS Algorithm 2.3 Limited Discrepancy Search (LDS) The original LDS algorithm was developed for searching a binary tree. As shown in [2], there are several alternative ways to modify LDS for a non-binary search tree. We define the discrepancy of node n (represented as dis(n)) as follows. The rank of node n (represented as rank(n)) is defined as the order amongits siblings, i.e., the rank of the best child node is 0, the second-best node is, and so on. By representingthe parent node as n p, dis(n) is defined as dis(n p )+rank(n), where the discrepancy of the root node dis(root) = 0. Figure 2 shows the discrepancies of nodes of a search tree in Figure. Figure 3 shows the pseudo code of the LDS algorithm. We store a list of nodes, which are siblings sorted by f, intostack. f max represents the highest price of the solutions found so far, and D max is the maximal number of allowed discrepancies. Initially, the stack contains a list of the root node, and f max is 0. 3 Experiments We ran a series of experiments for several problem settings used in previous works [,5] on a workstation (333 MHz Sun UltraSparc IIi with 52 MB) with a program written in C++. Due to space limitations, we only show the graphs of experiments on the followingtwo distributions used in [5]. Let M be the number of items and N be the, where bids are different from one another. The trends of the obtained results on other problem settings used in [] were also similar to the results described here. Random Distribution The number of items in each bid is randomly chosen from [,M], and items included in a bid are also randomly chosen. The price is randomly chosen from [0, ]. Uniform Distribution The number of items is set to a constant k (we set k = 3). The items and the price are chosen in the same manner as above. We generated 0 problem instances and calculated the average of them. To compare the results of LDS without re-calculation and the result of IDA* with re-calculation, we present () quality of obtained solution, i.e., the ratio of the
4 552 Yuko Sakurai et al. Average Number of Visited Nodes e+08 e+07 e () Solution Quality IDA* without re-calc. IDA* with re-calc (3) Visited Nodes Average Time Average Social Surplus IDA* without re-calc. IDA* with re-calc (2) Running Time 8 6 IDA* (4) Social Surplus LDS without re-calculation LDS with re-calculation Time (sec) (5) Anytime Performance Fig. 4. Random Distribution sum of the winningbids in LDS to the sum in IDA*, (2) runningtime, (3) number of visited nodes, and (4) social surplus, i.e., the sum of the winning bids, while varyingthe number of submitted bids, where we limit the maximal discrepancy to i [0, 2] (represented as Di) inlds. Random (Figures 4): As a comparison, we also show the results of IDA* without re-calculation. Figures 4 show the running time and the number of visited nodes of IDA* increase exponentially (note that the y-axis is logscaled), while those of LDS increase rather slowly. LDS with D2 remains better than 95% even for lage-scale problem instances. LDS can be considered an anytime algorithm. This property is desirable for winner determination problems. We illustrate (5) a comparison of the average anytime performance of LDS without re-calculation and with recalculation. We generated 0 instances where M = 400 and N = 500 and
5 An Efficient Approximate Algorithm for Winner Determination () Solution Quality Average Time Fig. 5. Uniform Distribution IDA* (2) Running Time gradually increased the maximal discrepancy (one-by-one from D0). This result shows, if we do not have enough time, it is reasonable to use the LDS algorithm without re-calculation to obtain a semi-optimal solution. Uniform (Figures 5): Figures 5 show the results where we set M = 00. IDA* can solve only problem instances where the is small within a reasonable amount of time. On the other hand, since LDS restricts the search space accordingto the maximal discrepancy, LDS can find about 95% of the optimal solution very quickly. 4 Conclusions Determiningthe winners in combinatorial auctions is a complicated constraint optimization problem. We have presented an approximate algorithm that is based on the LDS technique. Experimental evaluations performed on various problem sets showed that the LDS algorithm can find a semi-optimal solution (better than 95%) with a small amount of search effort (in about % of the runningtime) compared with the existingoptimal algorithm. References. Fujishima, Y., Leyton-Brown, K., and Shoham, Y.: Taming the Computational Complexity of Combinatorial Auctions: Optimal and Approximate Approaches. IJCAI-99 (999) 549, 550, Harvey,W.D.andGinsberg,M.L.:LimitedDiscrepancySearch.IJCAI-95 (995) 549, Korf, R. E.: Depth-first iterative deepening: an optimal admissible tree search. Artificial Intelligence 62 (993) Rothkopf, M. H., Pekeč, A., and Harstad, R. M.: Computationally Manageable Combinatorial Auctions. Management Science 44 (998) Sandholm, T.: An Algorithm for Optimal Winner Determination in Combinatorial Auction. IJCAI-99 (999) 549, 550, 55
Some Tractable Combinatorial Auctions
From: AAAI-00 Proceedings. Copyright 2000, AAAI (www.aaai.org). All rights reserved. Some Tractable Combinatorial Auctions Moshe Tennenholtz Faculty of Industrial Engineering and Management Technion, Israel
More informationWinner determination problem The auctioneer has a set of items, M = {1, 2,...,m}, to sell, and the buyers submit a set of bids, B = Introduction
From: AAAI-00 Proceedings. Copyright 2000, AAAI (www.aaai.org). All rights reserved. Improved Algorithms for Optimal Winner Determination in Combinatorial Auctions and Generalizations Abstract Combinatorial
More informationImproving Combinatorial Auctions for Multi-Robot Exploration
Improving Combinatorial Auctions for Multi-Robot Exploration Rodolfo C. Cavalcante Thiago F. Noronha Luiz Chaimowicz Abstract The use of multiple robots in exploration missions has attracted much attention
More informationEfficient memory-bounded search methods
Efficient memory-bounded search methods Mikhail Simin Arjang Fahim CSCE 580: Artificial Intelligence Fall 2011 Dr. Marco Voltorta Outline of The Presentation Motivations and Objectives Background - BFS
More informationHeuristic (Informed) Search
Heuristic (Informed) Search (Where we try to choose smartly) R&N: Chap., Sect..1 3 1 Search Algorithm #2 SEARCH#2 1. INSERT(initial-node,Open-List) 2. Repeat: a. If empty(open-list) then return failure
More informationChapter 14: Optimal Winner Determination Algorithms
Chapter 14: Optimal Winner Determination Algorithms Tuomas Sandholm 1 Introduction This chapter discusses optimal winner determination algorithms for combinatorial auctions (CAs). We say the auctioneer
More informationBest-First Search! Minimizing Space or Time!! RBFS! Save space, take more time!
Best-First Search! Minimizing Space or Time!! RBFS! Save space, take more time! RBFS-1 RBFS general properties! Similar to A* algorithm developed for heuristic search! RBFS-2 RBFS general properties 2!
More informationA Non-exact Approach and Experiment Studies on the Combinatorial Auction Problem
A Non-exact Approach and Experiment Studies on the Combinatorial Auction Problem Y. Guo 1, A. Lim 2, B. Rodrigues 3 and Y. Zhu 2 1 Department of Computer Science, National University of Singapore 3 Science
More informationArtificial Intelligence
Artificial Intelligence Informed Search and Exploration Chapter 4 (4.1 4.2) A General Search algorithm: Chapter 3: Search Strategies Task : Find a sequence of actions leading from the initial state to
More informationDFS. Depth-limited Search
DFS Completeness? No, fails in infinite depth spaces or spaces with loops Yes, assuming state space finite. Time complexity? O(b m ), terrible if m is much bigger than d. can do well if lots of goals Space
More informationWinner determination in combinatorial auctions with logic-based bidding languages Uckelman, J.D.; Endriss, U.
UvA-DARE (Digital Academic Repository) Winner determination in combinatorial auctions with logic-based bidding languages Uckelman, J.D.; Endriss, U. Published in: AAMAS 2008: 7th International Conference
More informationImpersonation-Based Mechanisms
Impersonation-Based Mechanisms Moshe Babaioff, Ron Lavi, and Elan Pavlov Abstract In this paper we present a general scheme to create mechanisms that approximate the social welfare in the presence of selfish
More informationNotes. Video Game AI: Lecture 5 Planning for Pathfinding. Lecture Overview. Knowledge vs Search. Jonathan Schaeffer this Friday
Notes Video Game AI: Lecture 5 Planning for Pathfinding Nathan Sturtevant COMP 3705 Jonathan Schaeffer this Friday Planning vs localization We cover planning today Localization is just mapping a real-valued
More informationAN INTEGER PROGRAMMING FORMULATION OF THE BID EVALUATION PROBLEM FOR COORDINATED TASKS
AN INTEGER PROGRAMMING FORMULATION OF THE BID EVALUATION PROBLEM FOR COORDINATED TASKS JOHN COLLINS AND MARIA GINI Abstract. We extend the IP models proposed by Nisan and Andersson for winner determination
More informationToday s s lecture. Lecture 3: Search - 2. Problem Solving by Search. Agent vs. Conventional AI View. Victor R. Lesser. CMPSCI 683 Fall 2004
Today s s lecture Search and Agents Material at the end of last lecture Lecture 3: Search - 2 Victor R. Lesser CMPSCI 683 Fall 2004 Continuation of Simple Search The use of background knowledge to accelerate
More informationTractable combinatorial auctions and b-matching
Artificial Intelligence 140 (2002) 231 243 www.elsevier.com/locate/artint Research Note Tractable combinatorial auctions and b-matching Moshe Tennenholtz 1 Faculty of Industrial Engineering and Management,
More informationAdvanced A* Improvements
Advanced A* Improvements 1 Iterative Deepening A* (IDA*) Idea: Reduce memory requirement of A* by applying cutoff on values of f Consistent heuristic function h Algorithm IDA*: 1. Initialize cutoff to
More informationAI: Week 2. Tom Henderson. Fall 2014 CS 5300
AI: Week 2 Tom Henderson Fall 2014 What s a Problem? Initial state Actions Transition model Goal Test Path Cost Does this apply to: Problem: Get A in CS5300 Solution: action sequence from initial to goal
More information3 SOLVING PROBLEMS BY SEARCHING
48 3 SOLVING PROBLEMS BY SEARCHING A goal-based agent aims at solving problems by performing actions that lead to desirable states Let us first consider the uninformed situation in which the agent is not
More informationMoving On. 10. Single-agent Search. Applications. Why Alpha-Beta First?
Moving On 10. Single-agent Search Jonathan Schaeffer jonathan@cs.ualberta.ca www.cs.ualberta.ca/~jonathan Two-player adversary search is nice, but not all interesting problems can be mapped to games Large
More informationComparative Study of RBFS & ARBFS Algorithm
IOSR Journal of Computer Engineering (IOSR-JCE) e-issn: 78-066, p- ISSN: 78-877Volume 0, Issue 5 (Mar. - Apr. 0), PP 05-0 Comparative Study of RBFS & ARBFS Algorithm Disha Sharma, Sanjay Kumar Dubey (Information
More informationBasic Search Algorithms
Basic Search Algorithms Tsan-sheng Hsu tshsu@iis.sinica.edu.tw http://www.iis.sinica.edu.tw/~tshsu 1 Abstract The complexities of various search algorithms are considered in terms of time, space, and cost
More informationMidterm Examination CS540-2: Introduction to Artificial Intelligence
Midterm Examination CS540-2: Introduction to Artificial Intelligence March 15, 2018 LAST NAME: FIRST NAME: Problem Score Max Score 1 12 2 13 3 9 4 11 5 8 6 13 7 9 8 16 9 9 Total 100 Question 1. [12] Search
More informationInformed (Heuristic) Search. Idea: be smart about what paths to try.
Informed (Heuristic) Search Idea: be smart about what paths to try. 1 Blind Search vs. Informed Search What s the difference? How do we formally specify this? A node is selected for expansion based on
More informationImproving the Efficiency of Depth-First Search by Cycle Elimination
Improving the Efficiency of Depth-First Search by Cycle Elimination John F. Dillenburg and Peter C. Nelson * Department of Electrical Engineering and Computer Science (M/C 154) University of Illinois Chicago,
More informationThis lecture. Lecture 6: Search 5. Other Time and Space Variations of A* Victor R. Lesser. RBFS - Recursive Best-First Search Algorithm
Lecture 6: Search 5 Victor R. Lesser CMPSCI 683 Fall 2010 This lecture Other Time and Space Variations of A* Finish off RBFS SMA* Anytime A* RTA* (maybe if have time) RBFS - Recursive Best-First Search
More informationInformed Search Methods
Informed Search Methods How can we improve searching strategy by using intelligence? Map example: Heuristic: Expand those nodes closest in as the crow flies distance to goal 8-puzzle: Heuristic: Expand
More informationAn Investigation of Representations of Combinatorial Auctions
An Investigation of Representations of Combinatorial Auctions David Loker and Kate Larson University of Waterloo 00 University Ave. W Waterloo, ON, Canada NL 3G {dloker,klarson}@cs.uwaterloo.ca ABSTRACT
More informationIntelligent Agents. Foundations of Artificial Intelligence. Problem-Solving as Search. A Simple Reflex Agent. Agent with Model and Internal State
Intelligent s Foundations of Artificial Intelligence Problem-Solving as Search S7 Fall 007 Thorsten Joachims : Anything that can be viewed as perceiving its environment through sensors and acting upon
More informationCombinatorial Auctions: A Survey by de Vries and Vohra
Combinatorial Auctions: A Survey by de Vries and Vohra Ashwin Ganesan EE228, Fall 2003 September 30, 2003 1 Combinatorial Auctions Problem N is the set of bidders, M is the set of objects b j (S) is the
More informationInformed search algorithms
Informed search algorithms This lecture topic Chapter 3.5-3.7 Next lecture topic Chapter 4.1-4.2 (Please read lecture topic material before and after each lecture on that topic) Outline Review limitations
More informationHomework 2: Multi-unit combinatorial auctions (due Nov. 7 before class)
CPS 590.1 - Linear and integer programming Homework 2: Multi-unit combinatorial auctions (due Nov. 7 before class) Please read the rules for assignments on the course web page. Contact Vince (conitzer@cs.duke.edu)
More informationAvailable from Deakin Research Online:
Deakin Research Online Deakin University s institutional research repository DDeakin Research Online Research Online This is the published version (version of record) of: Bai, H. and Zhang, Zili 2005,
More informationmywbut.com Informed Search Strategies-II
Informed Search Strategies-II 1 3.3 Iterative-Deepening A* 3.3.1 IDA* Algorithm Iterative deepening A* or IDA* is similar to iterative-deepening depth-first, but with the following modifications: The depth
More informationITCS 6150 Intelligent Systems. Lecture 5 Informed Searches
ITCS 6150 Intelligent Systems Lecture 5 Informed Searches Informed Searches We are informed (in some way) about future states and future paths We use this information to make better decisions about which
More informationTractable Combinatorial Auctions Via Graph Matching
Journal of Artificial Intelligence Research - (2007) p p- Submitted 05/07; published -/- Tractable Combinatorial Auctions Via Graph Matching Frank Guerin Department of Computing Science, University of
More informationLecture 9. Heuristic search, continued. CS-424 Gregory Dudek
Lecture 9 Heuristic search, continued A* revisited Reminder: with A* we want to find the best-cost (C ) path to the goal first. To do this, all we have to do is make sure our cost estimates are less than
More informationTrees : Part 1. Section 4.1. Theory and Terminology. A Tree? A Tree? Theory and Terminology. Theory and Terminology
Trees : Part Section. () (2) Preorder, Postorder and Levelorder Traversals Definition: A tree is a connected graph with no cycles Consequences: Between any two vertices, there is exactly one unique path
More informationArtificial Intelligence (part 4c) Strategies for State Space Search. (Informed..Heuristic search)
Artificial Intelligence (part 4c) Strategies for State Space Search (Informed..Heuristic search) Search Strategies (The Order..) Uninformed Search breadth-first depth-first iterative deepening uniform-cost
More informationLecture 5 Heuristics. Last Time: A* Search
CSE 473 Lecture 5 Heuristics CSE AI Faculty Last Time: A* Search Use an evaluation function f(n) for node n. f(n) = estimated total cost of path thru n to goal f(n) = g(n) + h(n) g(n) = cost so far to
More informationAsynchronous Weak-commitment Search for Solving Distributed Constraint Satisfaction Problems
International Conference on Principles and Practice of Constraint Programming 1995, pp.88 102 Asynchronous Weak-commitment Search for Solving Distributed Constraint Satisfaction Problems Makoto Yokoo NTT
More informationLearning techniques for Automatic Algorithm Portfolio Selection
Learning techniques for Automatic Algorithm Portfolio Selection Alessio Guerri and Michela Milano 1 Abstract. The purpose of this paper is to show that a well known machine learning technique based on
More informationSearch : Lecture 2. September 9, 2003
Search 6.825: Lecture 2 September 9, 2003 1 Problem-Solving Problems When your environment can be effectively modeled as having discrete states and actions deterministic, known world dynamics known initial
More informationmywbut.com Informed Search Strategies-I
Informed Search Strategies-I 1 3.1 Introduction We have outlined the different types of search strategies. In the earlier chapter we have looked at different blind search strategies. Uninformed search
More informationChapter 3: Solving Problems by Searching
Chapter 3: Solving Problems by Searching Prepared by: Dr. Ziad Kobti 1 Problem-Solving Agent Reflex agent -> base its actions on a direct mapping from states to actions. Cannot operate well in large environments
More informationArtificial Intelligence
Artificial Intelligence Dr. Malek Mouhoub Department of Computer Science University of Regina Fall 2005 Malek Mouhoub, CS820 Fall 2005 1 3. State-Space Search 3. State-Space Search Graph Theory Uninformed
More informationIterative deepening multiobjective A*
Iterative deepening multiobjective A* S. Harikumar, Shashi Kumar * Depurtment of Computer Science and Engineering, Indiun Institute of Technology, Delhi, New Delhi, India Received 19 January 1995; revised
More informationCrossword Puzzles as a Constraint Problem
Crossword Puzzles as a Constraint Problem Anbulagan and Adi Botea NICTA and Australian National University, Canberra, Australia {anbulagan,adi.botea}@nicta.com.au Abstract. We present new results in crossword
More informationAr#ficial)Intelligence!!
Introduc*on! Ar#ficial)Intelligence!! Roman Barták Department of Theoretical Computer Science and Mathematical Logic Uninformed (blind) search algorithms can find an (optimal) solution to the problem,
More informationCOMP9414: Artificial Intelligence Informed Search
COMP9, Wednesday March, 00 Informed Search COMP9: Artificial Intelligence Informed Search Wayne Wobcke Room J- wobcke@cse.unsw.edu.au Based on slides by Maurice Pagnucco Overview Heuristics Informed Search
More informationInteger Programming ISE 418. Lecture 7. Dr. Ted Ralphs
Integer Programming ISE 418 Lecture 7 Dr. Ted Ralphs ISE 418 Lecture 7 1 Reading for This Lecture Nemhauser and Wolsey Sections II.3.1, II.3.6, II.4.1, II.4.2, II.5.4 Wolsey Chapter 7 CCZ Chapter 1 Constraint
More informationInformed Search. CS 486/686 University of Waterloo May 10. cs486/686 Lecture Slides 2005 (c) K. Larson and P. Poupart
Informed Search CS 486/686 University of Waterloo May 0 Outline Using knowledge Heuristics Best-first search Greedy best-first search A* search Other variations of A* Back to heuristics 2 Recall from last
More informationTBBL: A Tree-Based Bidding Language for Iterative Combinatorial Exchanges
TL: Tree-ased idding Language for Iterative Combinatorial Exchanges The Harvard community has made this article openly available. Please share how this access benefits you. Your story matters Citation
More informationMonotonicity. Admissible Search: That finds the shortest path to the Goal. Monotonicity: local admissibility is called MONOTONICITY
Monotonicity Admissible Search: That finds the shortest path to the Goal Monotonicity: local admissibility is called MONOTONICITY This property ensures consistently minimal path to each state they encounter
More informationInformed search methods
Informed search methods Tuomas Sandholm Computer Science Department Carnegie Mellon University Read Section 3.5-3.7 of Russell and Norvig Informed Search Methods Heuristic = to find, to discover Heuristic
More informationBlind (Uninformed) Search (Where we systematically explore alternatives)
Blind (Uninformed) Search (Where we systematically explore alternatives) R&N: Chap. 3, Sect. 3.3 5 Slides from Jean-Claude Latombe at Stanford University (used with permission) Simple Problem-Solving-Agent
More informationHW#1 due today. HW#2 due Monday, 9/09/13, in class Continue reading Chapter 3
9-04-2013 Uninformed (blind) search algorithms Breadth-First Search (BFS) Uniform-Cost Search Depth-First Search (DFS) Depth-Limited Search Iterative Deepening Best-First Search HW#1 due today HW#2 due
More informationThe Distributed Constraint Satisfaction Problem: Formalization and Algorithms
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING, VOL. 10, NO. 5, SEPTEMBER/OCTOBER 1998 673 The Distributed Constraint Satisfaction Problem: Formalization and Algorithms Makoto Yokoo, Edmund H. Durfee,
More informationFinding Optimal Solutions to the Graph Partitioning Problem with Heuristic Search
Finding Optimal Solutions to the Graph Partitioning Problem with Heuristic Search Ariel Felner Department of Information Systems Engineering, Ben-Gurion University of the Negev Beer-Sheva, 85104, Israel
More informationRoute planning / Search Movement Group behavior Decision making
Game AI Where is the AI Route planning / Search Movement Group behavior Decision making General Search Algorithm Design Keep a pair of set of states: One, the set of states to explore, called the open
More informationA4B36ZUI - Introduction ARTIFICIAL INTELLIGENCE
A4B36ZUI - Introduction to ARTIFICIAL INTELLIGENCE https://cw.fel.cvut.cz/wiki/courses/a4b33zui/start Michal Pechoucek, Branislav Bosansky, Jiri Klema & Olga Stepankova Department of Computer Science Czech
More informationIntegrating Machine Learning in Parallel Heuristic Search
From: Proceedings of the Eleventh International FLAIRS Conference. Copyright 1998, AAAI (www.aaai.org). All rights reserved. Integrating Machine Learning in Parallel Heuristic Search R. Craig Varnell Stephen
More informationMinimal Preference Elicitation in Combinatorial Auctions
Minimal Preference Elicitation in Combinatorial Auctions Wolfram Conen XONAR GmbH Wodanstr. 7 42555 Velbert, Germany E-mail: conen@gmx.de Tuomas Sandholm Λ Carnegie Mellon University Computer Science Department
More informationAn Appropriate Search Algorithm for Finding Grid Resources
An Appropriate Search Algorithm for Finding Grid Resources Olusegun O. A. 1, Babatunde A. N. 2, Omotehinwa T. O. 3,Aremu D. R. 4, Balogun B. F. 5 1,4 Department of Computer Science University of Ilorin,
More informationBlind (Uninformed) Search (Where we systematically explore alternatives)
Blind (Uninformed) Search (Where we systematically explore alternatives) R&N: Chap. 3, Sect. 3.3 5 1 Simple Problem-Solving-Agent Agent Algorithm 1. s 0 sense/read initial state 2. GOAL? select/read goal
More informationLecture 2: Fun with Search. Rachel Greenstadt CS 510, October 5, 2017
Lecture 2: Fun with Search Rachel Greenstadt CS 510, October 5, 2017 Reminder! Project pre-proposals due tonight Overview Uninformed search BFS, DFS, Uniform-Cost, Graph-Search Informed search Heuristics,
More informationBreadth-first heuristic search. Paper by Rong Zhou, Eric A. Hansen Presentation by Salomé Simon
Breadth-first heuristic search Paper by Rong Zhou, Eric A. Hansen Presentation by Salomé Simon Breadth-first tree search 1 2 3 4 5 6 7 Used for search problems with uniform edge cost Prerequisite for presented
More informationCombinatorial Double Auction Winner Determination in Cloud Computing using Hybrid Genetic and Simulated Annealing Algorithm
Combinatorial Double Auction Winner Determination in Cloud Computing using Hybrid Genetic and Simulated Annealing Algorithm Ali Sadigh Yengi Kand, Ali Asghar Pourhai Kazem Department of Computer Engineering,
More informationDIT411/TIN175, Artificial Intelligence. Peter Ljunglöf. 23 January, 2018
DIT411/TIN175, Artificial Intelligence Chapters 3 4: More search algorithms CHAPTERS 3 4: MORE SEARCH ALGORITHMS DIT411/TIN175, Artificial Intelligence Peter Ljunglöf 23 January, 2018 1 TABLE OF CONTENTS
More informationTrees. CSE 373 Data Structures
Trees CSE 373 Data Structures Readings Reading Chapter 7 Trees 2 Why Do We Need Trees? Lists, Stacks, and Queues are linear relationships Information often contains hierarchical relationships File directories
More informationCOMP3702/7702 Artificial Intelligence Week2: Search (Russell & Norvig ch. 3)" Hanna Kurniawati"
COMP3702/7702 Artificial Intelligence Week2: Search (Russell & Norvig ch. 3)" Hanna Kurniawati" Last week" What is Artificial Intelligence?" Some history" Agent defined" The agent design problem" Search:
More informationARTIFICIAL INTELLIGENCE LECTURE 3. Ph. D. Lect. Horia Popa Andreescu rd year, semester 5
ARTIFICIAL INTELLIGENCE LECTURE 3 Ph. D. Lect. Horia Popa Andreescu 2012-2013 3 rd year, semester 5 The slides for this lecture are based (partially) on chapter 4 of the Stuart Russel Lecture Notes [R,
More informationApproximation Techniques for Utilitarian Mechanism Design
Approximation Techniques for Utilitarian Mechanism Design Department of Computer Science RWTH Aachen Germany joint work with Patrick Briest and Piotr Krysta 05/16/2006 1 Introduction to Utilitarian Mechanism
More informationCS 4700: Foundations of Artificial Intelligence
CS 4700: Foundations of Artificial Intelligence Bart Selman selman@cs.cornell.edu Module: Informed Search Readings R&N - Chapter 3: 3.5 and 3.6 Search Search strategies determined by choice of node (in
More informationSome Applications of Graph Bandwidth to Constraint Satisfaction Problems
Some Applications of Graph Bandwidth to Constraint Satisfaction Problems Ramin Zabih Computer Science Department Stanford University Stanford, California 94305 Abstract Bandwidth is a fundamental concept
More informationA.I.: Informed Search Algorithms. Chapter III: Part Deux
A.I.: Informed Search Algorithms Chapter III: Part Deux Best-first search Greedy best-first search A * search Heuristics Outline Overview Informed Search: uses problem-specific knowledge. General approach:
More informationHEURISTIC SEARCH. 4.3 Using Heuristics in Games 4.4 Complexity Issues 4.5 Epilogue and References 4.6 Exercises
4 HEURISTIC SEARCH Slide 4.1 4.0 Introduction 4.1 An Algorithm for Heuristic Search 4.2 Admissibility, Monotonicity, and Informedness 4.3 Using Heuristics in Games 4.4 Complexity Issues 4.5 Epilogue and
More informationCS 331: Artificial Intelligence Informed Search. Informed Search
CS 331: Artificial Intelligence Informed Search 1 Informed Search How can we make search smarter? Use problem-specific knowledge beyond the definition of the problem itself Specifically, incorporate knowledge
More informationInformed Search Algorithms
Informed Search Algorithms CITS3001 Algorithms, Agents and Artificial Intelligence Tim French School of Computer Science and Software Engineering The University of Western Australia 2017, Semester 2 Introduction
More informationHeuris'c Search. Reading note: Chapter 4 covers heuristic search.
Heuris'c Search Reading note: Chapter 4 covers heuristic search. Credits: Slides in this deck are drawn from or inspired by a multitude of sources including: Shaul Markovitch Jurgen Strum Sheila McIlraith
More informationDistributed Constraint Satisfaction Algorithm for Complex Local Problems
Third International Conference on Multiagent Systems (ICMAS-98), pp.372 379, 1998 Distributed Constraint Satisfaction Algorithm for Complex Local Problems Makoto Yokoo NTT Communication Science Laboratories
More informationA Re-examination of Limited Discrepancy Search
A Re-examination of Limited Discrepancy Search W. Ken Jackson, Morten Irgens, and William S. Havens Intelligent Systems Lab, Centre for Systems Science Simon Fraser University Burnaby, B.C., CANADA V5A
More informationCS 331: Artificial Intelligence Informed Search. Informed Search
CS 331: Artificial Intelligence Informed Search 1 Informed Search How can we make search smarter? Use problem-specific knowledge beyond the definition of the problem itself Specifically, incorporate knowledge
More information1 Introduction. 2 Iterative-Deepening A* 3 Test domains
From: AAAI Technical Report SS-93-04. Compilation copyright 1993, AAAI (www.aaai.org). All rights reserved. Fast Information Distribution for Massively Parallel IDA* Search Diane J. Cook Department of
More informationData Abstractions. National Chiao Tung University Chun-Jen Tsai 05/23/2012
Data Abstractions National Chiao Tung University Chun-Jen Tsai 05/23/2012 Concept of Data Structures How do we store some conceptual structure in a linear memory? For example, an organization chart: 2/32
More informationRichard E. Korf. June 27, Abstract. divide them into two subsets, so that the sum of the numbers in
A Complete Anytime Algorithm for Number Partitioning Richard E. Korf Computer Science Department University of California, Los Angeles Los Angeles, Ca. 90095 korf@cs.ucla.edu June 27, 1997 Abstract Given
More informationSemi-Independent Partitioning: A Method for Bounding the Solution to COP s
Semi-Independent Partitioning: A Method for Bounding the Solution to COP s David Larkin University of California, Irvine Abstract. In this paper we introduce a new method for bounding the solution to constraint
More informationCS 380/480 Foundations of Artificial Intelligence Winter 2007 Assignment 2 Solutions to Selected Problems
CS 380/480 Foundations of Artificial Intelligence Winter 2007 Assignment 2 Solutions to Selected Problems 1. Search trees for the state-space graph given below: We only show the search trees corresponding
More informationTime complexity of iterative-deepening-a
Artificial Intelligence 129 (2001) 199 218 Time complexity of iterative-deepening-a Richard E. Korf a,, Michael Reid b, Stefan Edelkamp c a Computer Science Department, University of California, Los Angeles,
More informationHeuristic Search. CPSC 470/570 Artificial Intelligence Brian Scassellati
Heuristic Search CPSC 470/570 Artificial Intelligence Brian Scassellati Goal Formulation 200 Denver 300 200 200 Chicago 150 200 Boston 50 1200 210 75 320 255 Key West New York Well-defined function that
More informationMulti-Way Number Partitioning
Proceedings of the Twenty-First International Joint Conference on Artificial Intelligence (IJCAI-09) Multi-Way Number Partitioning Richard E. Korf Computer Science Department University of California,
More informationArtificial Intelligence
Artificial Intelligence Search Marc Toussaint University of Stuttgart Winter 2015/16 (slides based on Stuart Russell s AI course) Outline Problem formulation & examples Basic search algorithms 2/100 Example:
More informationChapters 3-5 Problem Solving using Search
CSEP 573 Chapters 3-5 Problem Solving using Search First, they do an on-line search CSE AI Faculty Example: The 8-puzzle Example: The 8-puzzle 1 2 3 8 4 7 6 5 1 2 3 4 5 6 7 8 2 Example: Route Planning
More informationInformed search strategies (Section ) Source: Fotolia
Informed search strategies (Section 3.5-3.6) Source: Fotolia Review: Tree search Initialize the frontier using the starting state While the frontier is not empty Choose a frontier node to expand according
More informationCS 520: Introduction to Artificial Intelligence. Lectures on Search
CS 520: Introduction to Artificial Intelligence Prof. Louis Steinberg Lecture : uninformed search uninformed search Review Lectures on Search Formulation of search problems. State Spaces Uninformed (blind)
More informationState Space Search. Many problems can be represented as a set of states and a set of rules of how one state is transformed to another.
State Space Search Many problems can be represented as a set of states and a set of rules of how one state is transformed to another. The problem is how to reach a particular goal state, starting from
More informationCOMP9414: Artificial Intelligence Informed Search
COMP9, Monday 9 March, 0 Informed Search COMP9: Artificial Intelligence Informed Search Wayne Wobcke Room J- wobcke@cse.unsw.edu.au Based on slides by Maurice Pagnucco Overview Heuristics Informed Search
More informationHeuristic Search in Bounded-depth Trees: Best-Leaf-First Search
Heuristic Search in Bounded-depth Trees: Best-Leaf-First Search Wheeler Ruml Division of Engineering and Applied Sciences Harvard University Cambridge, MA 02138 ruml@eecs.harvard.edu Abstract Many combinatorial
More informationChapter 3: Informed Search and Exploration. Dr. Daisy Tang
Chapter 3: Informed Search and Exploration Dr. Daisy Tang Informed Search Definition: Use problem-specific knowledge beyond the definition of the problem itself Can find solutions more efficiently Best-first
More informationCS 4700: Foundations of Artificial Intelligence
CS 4700: Foundations of Artificial Intelligence Bart Selman selman@cs.cornell.edu Informed Search Readings R&N - Chapter 3: 3.5 and 3.6 Search Search strategies determined by choice of node (in queue)
More information